Psychometric properties Re-evaluation of the Persian version of Manchester Driving Behavior Questionnaire

Background and aims: Road traffic accidents are considered a major threat to public health worldwide and are usually referred as “hidden epidemic”. In low and middle-income countries, however, worries are more severe. For example, a national study in Iran found that traffic accidents ranked first in the number of years of life lost due to premature death in Iran. Various studies have divided the factors affecting the occurrence of traffic accidents into three categories: human, vehicle, and environmental factors. The human factor has a decisive role in the occurrence of accidents. In a way that based on studies, this role in traffic accidents could be 70 to 90%. It is clear that study in this area, like other research, requires data collection. Since archival information about drivers' behavior is often incomplete and difficult to obtain, researchers often have to evaluate driving behaviors using self-report tools. One of the foremost common selfreporting instruments used in research focused on driver behavior is the Manchester Driving Behavior Questionnaire (MDBQ). The only complete Persian version of MDBQ has been published by Oreyzi et al. (2010). The authors of the present study had initially intended to use a Persian MDBQ questionnaire to conduct a field study. But while comparing this questionnaire with the original one (in English), they found several shortcomings in the Persian version. For example, it can be mentioned the lack of attention to the difference between regular driving of the road in the UK (on the left side) and in Iran (on the right side). Since the accuracy of data collection tools is vital in research, the authors intended to revise the Persian version of this widely used questionnaire. The purpose of this study was to re-evaluate the psychometric properties of MDBQ. Methods: The original MDBQ version developed by Reason (1990) with 50 items responding in a five-point Likert scale (never = 1 to always = 5). The items are divided into four categories according to the type of behavior including slip, deliberate violation, mistake, and unintentional violation. To evaluate the psychometric properties of the Persian version of MDBQ, the linguistic, face, and construct validity were check, as well as, instrument reliability. For assessing the linguistic validity of the questionnaire, the original version of the DBQ questionnaire was translated by English-speaking experts, one of whom had more than ten years of driving experience in European countries. In this way, it was possible to implement traffic signs and regulations in the country of origin and Iran (localization of items). This was followed by other linguistic validity steps. To investigate the qualitative face validity of the questionnaire, the questionnaire was given to 15 drivers to express their views on each of the questionnaire items in terms of difficulty, appropriateness, ambiguity, relevance, and ambiguity of the instrument items. To study the quantitative face validity of the items (impact factor), the questionnaire was given to 14 members of the target group (drivers) and asked to rate the importance of each item in the questionnaire in a five-point Likert scale (from not important at all = 1 to very important = 5). The criterion for accepting each item is the score of the impact factor, which should not be less than 1.5. Content validity was studied using the content validity ratio (CVR) and content validity index (CVI). For this purpose, a questionnaire was provided to 10 members of the expert panel (expert on safety and occupational health and health education) who were fully acquainted with the content validity assessment. After making the necessary changes based on the expert panel opinions, the mentioned indicators were calculated. To evaluate the CVR (necessity of each item), the group of experts was asked to classify each item according to the three-point Likert scale (essential = 1, useful but not necessary = 2 and not necessary = 3). Then the content validity ratio was calculated. The CVR acceptance criterion based on the number of experts (n = 10) was 0.62 and the minimum CVI acceptance was 0.79. Keywords